Impact of prophylactic fundoplication on survival without growth disorders in left congenital diaphragmatic hernia requiring a patch repair

(1)

Impact of Prophylactic Fundoplication on Survival without Growth Disorder in Left Congenital Diaphragmatic Hernia Requiring a Patch Repair

Anne Dariel, MD, Jean-Christophe Roze´, MD, PhD, Hugues Piloquet, MD, Guillaume Podevin, MD, PhD, and the French CDH Study Group*

Growth impairment is frequent in surviving newborns with congenital diaphragmatic hernia requiring a patch repair.

This multicenter retrospective study included 57 newborns and showed a significant relationship between prophylactic fundoplication performed during initial diaphragmatic repair and survival without disordered growth, after adjustment for propensity score (adjusted OR 4.7 [1.2–18.5];P= .03).(J Pediatr 2010;-:---)

C

ongenital diaphragmatic hernia requiring a patch repair represents a challenge for medical-surgical teams.

Nutritional morbidity becomes one of the major chronic concerns for survivors^1,2and is mainly caused by the development of severe gastroesophageal reflux, which is reported in up to 60% of cases¹and frequently requires fundoplication during the first months of life.³ We investigated the hypothesis that a prophylactic fundoplication performed during initial diaphragmatic repair can limit nutritional morbidity.

Methods

Between December 1994 and June 2005, 61 cases of left-side congenital diaphragmatic hernia requiring a patch repair were treated in 8 French pediatric surgery units. Four cases were excluded from this study for severe prematurity (gesta- tional age < 30 weeks). Fifty-seven congenital diaphragmatic hernias requiring a patch repair were included retrospec- tively. The indication for patch repair was that direct closure was impossible because of diaphragmatic aplasia or severe hypoplasia.

To detect growth disorder, weight-for-height and height- for-age indicators were calculated according to child growth charts and tables from the World Health Organization. For premature infants, we used corrected age. Growth disorder was defined by weight-for-height and height-for-age ratios

< 1.5 standard deviations (SD), because a 2 SD threshold would have eliminated too many cases, and 1 SD would not have been discriminatory.

Because this study was retrospective and multicenter, we used the propensity score to reduce bias.⁴ The propensity score is defined as a conditional probability between 0 and 1 that a patient will be treated, on the basis of an observed group of covariates. The score is then used as if it were the only con- founding covariate to collapse the different covariates between the groups and then reduce bias. A logistic regression model was developed to determine the propensity score for prophylactic fundoplication. This model included covariates indi- cated inTable I. The 8 centers were divided into 3 groups

for the statistical analysis: centers 1 and 2 that treated more than 15 congenital diaphragmatic hernias requiring a patch repair, and 6 other centers that treated less than 15 cases each. Patients with missing data were excluded from multivariable analysis. Three logistic regression models were used to estimate adjusted odds ratios, with 95% confidence intervals, of survival without growth disorder, and survival without growth disorder and delayed fundoplication (performed for gastroesophageal reflux disease during a second surgical time): (1) no adjustment; (2) adjustment for centers; and (3) adjustment for propensity score. All analyses were performed with SPSS 15.0 (SPSS, Chicago, Illinois). Continuous values were expressed as medians (range values). The Mann-Whitney U test and thec²test, if necessary, were used to compare population characteristics.

Results

The population involved in our study is summarized inTable I. Prophylactic fundoplication was performed in 34 cases and not performed in 23. Antireflux techniques were chosen according to surgical habits and anatomic findings:

complete fundoplication (Nissen procedure) (n = 14) or partial fundoplication (Toupet procedure or ventral semi- valve) (n = 20). No complications specifically related to prophylactic fundoplication were noted, but fundoplication was redone in 4 cases. Parenteral nutrition was used in all 57 cases without significant difference in duration: 37 (range 11-105) and 36 (range 9-145) days (P = .73) in patients with and without prophylactic fundoplication respectively. A gastric tube was added in 14 cases and 10, respectively (P = .85). For the 23 cases without prophylactic fundoplication, delayed fundoplication was needed in 9 and performed at a median age of 3.25 months

From the Departments of Pediatric Surgery (A.D., G.P.) and Pediatrics (H.R.), and the Neonatal Intensive Care Unit (J-C.R.), University Hospital, and INSERM CIC 004, Nantes, France

*List of members of the French CDH Study Group is available atwww.jpeds.com (Appendix).

The authors declare no conflicts of interest.

0022-3476/$ - see front matter. Copyrightª2010 Mosby Inc.

SD Standard deviations

1

(2)

(range 2-8). Among these 9 infants, 6 were alive at 3 years of age, and in 5/6 infants we observed catch-up growth.

Forty-three infants survived, 29 with and 14 without prophylactic fundoplication (P= .04); 4/26 and 3/13 presented growth failure among 39 infants measured at 6 month (P= .36), 1/26 versus 5/13 at 1 year (P = .02) and 9/29 versus 11/14 (P= .04) at least once during follow-up. Overall, 20/34 and 5/23 survived without growth disorder at the end of follow- up (P = .02). Median follow-up was 5.0 (range 2.0-12.5) and 4.3 (range 3.0-7.2) years (P= .45), respectively.

Fourteen patients died, 5 with and 9 without prophylactic fundoplications. Median age at death was 30 (range 10-214) and 27 (range 3-457) days (P = .29), respectively. Eight patients died before 1 month because of severe pulmonary hypoplasia, 3 with and 5 without prophylactic fundoplication. Six patients died later because of worsening of pulmonary hypertension; growth disorders were present in 1 of the 2 cases with prophylactic fundoplication and in 3 of the 4 cases without prophylactic fundoplication.

Propensity Score

The propensity score was calculated for 56 of the 57 newborns included in the study. As noted in Table I, the choice of fundoplication during primary diaphragmatic repair was not influenced by sex, intrathoracic liver, or preoperative extracorporeal membrane oxygenation. Only centers (particularly center 2) and birth weight above 2500 g significantly increased or tended to increase, respectively, the probability of undergoing prophylactic fundoplication.

Association between Primary Outcome and Survival without Growth Disorder

In each model, prophylactic fundoplication was significantly associated with survival without growth disorder and survival without growth disorder and delayed fundoplication (Table II). Survival per se was not significantly associated with prophylactic fundoplication.

Discussion

In this multicenter retrospective study, performing prophylactic fundoplication during initial diaphragmatic repair is

associated with survival without disordered growth after adjustment with a propensity score, which is a way of asserting that the effect of a treatment is compared only between patients who are equally likely to receive it.⁴

Gastroesophageal reflux is a well-known complication in survivors of congenital diaphragmatic hernia that increases growth disorders and oral aversion. It has been reported in 30% to 81% of long-term survivors.³A patient with a congenital diaphragmatic hernia requiring a patch repair is likely to have development of gastroesophageal reflux needing fundoplication.^1,2Antireflux surgery is performed most commonly in the year after congenital diaphragmatic hernia repair.^5,6 Prophylactic fundoplication added no additional risk in this cohort.

Growth retardation is a common chronic concern in these patients with congenital diaphragmatic hernia, particularly in those requiring a patch repair: 55% of patients were below the 25th percentile in the first year of life in Murato’s report,² 40% had a weight below the fifth percentile at 1 year as reported by Lund et al,⁷and 21% had a reduced weight/height ratio at 2 years of age as reported by Van Meurs et al.⁸Early fundoplication improved long-term weight gain.⁹Before this study, the potential benefit of prophylactic fundoplication had not been evaluated, except in neonates with congenital heart diseases.

The retrospective nature of this study is a limiting factor.

Nevertheless when we adjusted the propensity score, we Table I. Associations between variables included in the calculation of propensity score and prophylactic fundoplication

Covariates

Infants with prophylactic fundoplication (n = 34)

Infants without prophylactic

fundoplication (n = 23) aOR 95% CI Pvalue

Male 18 (52.9%) 16 (69.6%) 1.0 (0.2-4.1) .991

Birth weight (<2500 g) 2 (5.9%) 4 (17.4%) 0.1 (0.0- 1.0) .051

Centers

Other Centers 11 (32.4%) 14 (60.9%) 1

Center 1 11 (32.4%) 6 (26.1%) 2.5 (0.4-16.7) .331

Center 2 12 (35.3%) 3 (13.0%) 11.3 (1.6-79.1) .015

Age at surgery (day)) 5 [2-10] 4 [2-6] 1.3 (1.0-1.7) .061

Intrathoracic liver 18 (52.9%) 14 (60.9%) 1.0 (0.1-11.8) .992

Associated malformations 8 (23.5%) 9 (39.1%) 0.2 (0.0-1.3) .101

Pre-operative ECMO 7 (20.6%) 4 (17.4%) 1.8 (0.2-15.6) .574

aOR,Adjusted odds ratio.

Data are presented as n (%) or median [interquartile].

Table II. Association between prophylactic fundoplication and survival without growth disorders, and survival without growth disorders and delayed fundoplication

OR 95% CI Pvalue

Survival without growth disorders

No adjustment 4.0 (1.3-12.8) .018

Adjustment for center 5.6 (1.6-20.6) .009

Adjustment for propensity score 4.7 (1.2-18.5) .028 Survival without growth disorders and

delayed fundoplication

No adjustment 9.5 (2.4-38.3) .002

Adjustment for center 13.1 (2.9-60.6) .001

Adjustment for propensity score 15.9 (2.8-88.3) .002 OR,Odds ratio.

T

^HE

J

^{OURNAL OF}

P

^EDIATRICS

www.jpeds.com

Vol.-, No.-

2 Dariel et al

ARTICLE IN PRESS

(3)

adjusted 7 covariates to reflect a center’s practice, including nutritional support practice. So the observed association seems to be not biased, although it should be confirmed by a randomized prospective study. We suggest that prophylactic fundoplication can prevent growth disorder in infants with left congenital diaphragmatic hernia requiring a patch repair. n

Submitted for publication Sep 22, 2009; last revision received Mar 31, 2010;

accepted Jun 7, 2010.

Reprint requests: Anne Dariel, MD, Hopital de la Mere et de l’Enfant, Service de Neonatalogie, 9 Quai Moncousu, 44099 Nantes Cedex, France.

E-mail:jcroze@chu-nantes.fr.

References

1.Su W, Berry M, Puligandla PS, Aspirot A, Flageole H, Laberge JM. Predic- tors of gastroesophageal reflux in neonates with congenital diaphragmatic hernia. J Pediatr Surg 2007;42:1639-43.

2.Muratore CS, Utter S, Jaksic T, Lund DP, Wilson JM. Nutritional morbidity in survivors of congenital diaphragmatic hernia. J Pediatr Surg 2001;

36:1171-6.

3.Koivusalo AI, Pakarinen MP, Lindahl HG, Rintala RJ. The cumulative in- cidence of significant gastroesophageal reflux in patients with congenital diaphragmatic hernia—a systematic clinical, pH-metric and endoscopic follow-up study. J Pediatr Surg 2008;43:279-82.

4.D’Agostino RB. Propensity score methods for bias reduction in the com- parison of a treatment to a non-randomized control group. Statist Med 1998;17:2265-81.

5.Diamond IR, Mah K, Kim PC, Bohn D, Gerstle JT, Wales PW. Predicting the need for fundoplicature at the time of congenital diaphragmatic hernia repair. J Pediatr Surg 2007;42:1066-70.

6.Rudolph CD, Mazur LJ, Liptak GS, Baker RD, Boyle JT, Colletti RB, et al.

Guidelines for evaluation and treatment of gastroesophageal reflux in infants and children: recommendations of the North American Society for Pediatric Gastroenterology and Nutrition. J Pediatr Gastroenterol Nutr 2001;32(Suppl 2):S1-31.

7.Lund DP, Mitchell J, Kharasch V, Quigley S, Kuehn M, Wilson JM. Con- genital diaphragmatic hernia: the hidden morbidity. J Pediatr Surg 1994;

29:258-62.

8.Van Meurs KP, Robbins ST, Reed VL, Karr SS, Wagner AE, Glass P, et al.

Congenital diaphragmatic hernia: long-term outcome in neonates treated with extracorporeal membrane oxygenation. J Pediatr Surg 1993;122:893-9.

9.Cribbs RK, Heiss KF, Clabby ML, Wulkan ML. Gastric fundoplicature is effective in promoting weight gain in children with severe congenital heart defects. J Pediatr Surg 2008;43:283-9.

-2010

CLINICAL AND LABORATORY OBSERVATIONS

Impact of Prophylactic Fundoplication on Survival without Growth Disorder in Left Congenital Diaphragmatic Hernia Requiring a Patch Repair

3

ARTICLE IN PRESS

(4)

Appendix

The following teams from the French CDH Study Group par- ticipating in this work:

E. Carricaburu, MD, Department of Pediatric Surgery, Robert Debre´ Hospital, AP-HP, Paris; C. Chamond, MD, Department of Pediatric Surgery, Saint-Vincent de Paul Hos- pital, AP-HP, Paris; P. de Lagausie, MD, PhD, Department of

Pediatric Surgery, University Hospital Timone Enfant, Mar- seille; M. Larroquet, MD, Department of Pediatric Surgery, Armand Trousseau Hospital, AP-HP, Paris; G. Levard, MD, PhD, Department of Pediatric Surgery, University Hos- pital la Mile´trie, Poitiers; R. Sfeir, MD, Department of Pedi- atric Surgery, University Hospital Jeanne de Flandre, Lille;

and D. Weil, MD, Department of Pediatric Surgery, Centre Hospitalier, Le Mans.

T

^HE

J

^{OURNAL OF}

P

^EDIATRICS

www.jpeds.com

Vol.-, No.-

3.e1 Dariel et al